aboutsummaryrefslogtreecommitdiff
path: root/wqflask
diff options
context:
space:
mode:
authorPjotr Prins2015-04-18 10:58:26 +0000
committerPjotr Prins2015-04-18 10:58:26 +0000
commit8706319923b3830a4d8cd63fd9a3f6b9a2b04563 (patch)
tree9fd446c188002fbc16f9f0d6090aca99def0844e /wqflask
parent02660b9406a97943d4c33946250fc3f08b80c556 (diff)
downloadgenenetwork2-8706319923b3830a4d8cd63fd9a3f6b9a2b04563.tar.gz
Fix NA to float tranforms
Diffstat (limited to 'wqflask')
-rw-r--r--wqflask/wqflask/my_pylmm/pyLMM/lmm.py37
1 files changed, 27 insertions, 10 deletions
diff --git a/wqflask/wqflask/my_pylmm/pyLMM/lmm.py b/wqflask/wqflask/my_pylmm/pyLMM/lmm.py
index 618f8332..5b06c9ae 100644
--- a/wqflask/wqflask/my_pylmm/pyLMM/lmm.py
+++ b/wqflask/wqflask/my_pylmm/pyLMM/lmm.py
@@ -856,30 +856,47 @@ def gwas_with_redis(key,species,new_code=True):
def narray(t):
info("Type is "+t)
- v = params[t]
+ v = params.get(t)
if v is not None:
- v = np.array(v).astype(np.float)
- print(v)
+ # Note input values can be array of string or float
+ v1 = [x if x != 'NA' else 'nan' for x in v]
+ v = np.array(v1).astype(np.float)
return v
- def narrayT(t):
- m = narray(t)
+ def marray(t):
+ info("Type is "+t)
+ v = params.get(t)
+ if v is not None:
+ m = []
+ for r in v:
+ # Note input values can be array of string or float
+ r1 = [x if x != 'NA' else 'nan' for x in r]
+ m.append(np.array(r1).astype(np.float))
+ return np.array(m)
+ return np.array(v)
+
+ def marrayT(t):
+ m = marray(t)
if m is not None:
return m.T
return m
# We are transposing before we enter run_gwas - this should happen on the webserver
# side (or when reading data from file)
- print(params)
- k = narray('kinship_matrix')
- g = narrayT('genotype_matrix')
+ k = marray('kinship_matrix')
+ g = marrayT('genotype_matrix')
+ mprint("geno",g)
y = narray('pheno_vector')
n = len(y)
- m = params['num_genotypes']
- ps,ts = run_gwas(species,n,m,k,y,g,narray('covariate_matrix'),params['restricted_max_likelihood'],params['refit'],params['input_file_name'],new_code)
+ m = params.get('num_genotypes')
+ if m is None:
+ m = g.shape[0]
+ info("m=%d,n=%d" % (m,n))
+ ps,ts = run_gwas(species,n,m,k,y,g,narray('covariate_matrix'),params['restricted_max_likelihood'],params['refit'],params.get('input_file_name'),new_code)
results_key = "pylmm:results:" + params['temp_uuid']
+ # fatal(results_key)
json_results = json.dumps(dict(p_values = ps,
t_stats = ts))